Toward Earth system modeling with resolved clouds and ocean submesoscales on heterogeneous many-core HPCs

Zhang, Shaoqing; Xu, Shiming; Fu, Haohuan; Wu, Lijun; Liu, Zhao; Gao, Yang; Zhao, Chun; Wan, Wubing; Wan, Lingfeng; Lu, Haitian; Li, Chenling; Liu, Yanfei; Lv, Xiaojing; Xie, Jin; Yu, Yangyang; Gu, Jiyou; Wang, Xuantong; Zhang, Yan; Ning, Chenhui; Fei, Yunlong; Guo, Xinyun; Wang, Zhaoying; Wang, Xue; Wang, Zhenming; Qu, Binglin; Li, Mingkui; Zhao, Haoran; Jiang, Yingjing; Yang, Guang‐Fu; Wang, Hong; An, Hong; Zhang, Xin; Ma, Wentao; Yu, Fujiang; Wang, Yuqing; Shen, Xueshun

doi:10.1093/nsr/nwad069

Cited by 10 publications

(3 citation statements)

References 88 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Concurrent execution of multiple components may induce numerical instability, which can be caused by the following conditions: inconsistent coastal line of atmosphere and ocean models, coupling data errors, imbalance of physical processes at the components interface, etc. The development of numerical instability due to these errors and the importance of correctly coupling the components is shown in [2,26,35,74]. To understand the feedbacks between components of…”

Section: Results Of Numerical Simulationsmentioning

confidence: 99%

The Parallel Performance of SLNE Atmosphere-Ocean-Sea Ice Coupled Model

2023

JSFI

View full text Add to dashboard Cite

The paper presents the first version of SLNE coupled model. SL and NE here are the first two letters from SLAV (Semi-Lagrangian, based on Absolute Vorticity equation) atmospheric model and NEMO (Nucleus for European Modelling of the Ocean) ocean model that have been coupled using OASIS3-MCT software. SLAV uses 0.9 • x0.72 • regular lat-lon grid with 96 vertical levels. NEMO incorporates SI3 sea ice model. Both of them use the same ORCA025 tripolar grid. Flux adjustments to correct inconsistencies at the interface between coupled atmosphereocean models have not been applied in SLNE. The model design and coupling particularities are described here in detail. A series of numerical experiments with SLNE model were performed to measure its parallel performance. We also investigated the scalability of SLNE model and its components in terms of simulation speed. Based on these results, an optimum configurations of SLNE were identified. It was found that the coupled model showed scaling efficiency of about 85% on 4000 computational cores of Cray XC40-LC in comparison to the SLNE configuration running on 224 cores. Simulations with lead times ranging from a few days to several years showed that there are no significant systematic errors in the coupled model.

show abstract

Section: Results Of Numerical Simulationsmentioning

confidence: 99%

The Parallel Performance of SLNE Atmosphere-Ocean-Sea Ice Coupled Model

2023

JSFI

View full text Add to dashboard Cite

show abstract

“…The intense submesoscale air‐sea interaction has profound implications for climate and weather projections. In the presence of submesoscale fronts, the ocean could release more heat and moisture into the atmosphere, which drives stronger local convection systems, including vertical motions, cloud formation and precipitation, that are still challenging to be represented in the mesoscale‐resolving climate models (Chang et al., 2020; Small et al., 2019; Yeager et al., 2023; Zhang et al., 2023). Besides, the energetic atmospheric frontal activities and the intensity of storm tracks are fueled by the air‐sea interactions in frontal regions (Hirata & Nonaka, 2021; Kuwano‐Yoshida & Minobe, 2017; O’Reilly & Czaja, 2015; O’Reilly et al., 2017; Parfitt & Czaja, 2016; Parfitt et al., 2016; Reeder et al., 2021).…”

Section: Summary and Discussionmentioning

confidence: 99%

Observations Reveal Intense Air‐Sea Exchanges Over Submesoscale Ocean Front

Yang,

Chen,

Sun

et al. 2024

Geophysical Research Letters

Self Cite

View full text Add to dashboard Cite

Air‐sea exchanges across oceanic fronts are critical in powering cloud formation, precipitation, and atmospheric storms. Oceanic submesoscale fronts of scales 1–10 km are characterized by strong sea surface temperature (SST) gradients. However, it remains elusive how submesoscale fronts affect the overlying atmosphere due to a lack of high‐resolution observations or models. Based on rare high‐resolution in situ observations in the Kuroshio Extension region, we quantify the air‐sea exchanges across an oceanic submesoscale front. The cross‐front SST and turbulent heat flux gradients reaches 2.4°C/km and 47 W/m2/km, respectively, far stronger than that typically found in mesoscale‐resolving products. The stronger SST gradient drives substantially stronger air‐sea fluxes and vertical mixing than mesoscale fronts, enhancing cloud formations. The intense air‐sea exchanges across submesoscale fronts are confirmed in idealized model simulations, but not resolved in mesoscale‐resolving climate models. Our finding provides essential knowledge for improving simulations of cloud formation, precipitation, and storms in climate models.

show abstract

“…Fortunately, models that meet these requirements have just become available, such as the new ICON atmosphere-ocean model developed at the Max Planck Institute for Meteorology that has a horizontal resolution of 5 km ([ 87 ], see [ 88 ] for a similar, recent example). The ICON simulation was exploited by Gutjahr et al .…”

Section: …And Where Rapid Tried But Has Not Delivered Yetmentioning

confidence: 99%

From theory to RAPID AMOC observations: a personal voyage of discovery

Marotzke

2023

Phil. Trans. R. Soc. A.

View full text Add to dashboard Cite

I review the history of ideas that have led to the establishment of the RAPID monitoring system for the Atlantic Meridional Overturning Circulation (AMOC) at 26.5° N. This history is closely connected to important events in my personal career. Starting from early and largely unsuccessful attempts at formulating a dynamically consistent force balance for the AMOC, I made theoretical progress by separately predicting the density at the eastern and western boundaries and by invoking exact geostrophic balance throughout, including the western boundary current. A remarkable confluence of individuals and ideas then enabled the establishment of the RAPID array, at its core based on monitoring boundary densities and on geostrophy. The RAPID results, such as the surprisingly large sub-seasonal variability, have encouraged AMOC monitoring approaches at other latitudes. I finish by pointing at two theoretical concepts—first, acknowledging the difference between convective mixing and sinking and, second, considering the advective rather than wave propagation of density perturbations in the deep western boundary current—that, together with continued observations and newly available global coupled simulations at very high resolution, should substantially improve our understanding of the causes of AMOC variability. This article is part of a discussion meeting issue 'Atlantic overturning: new observations and challenges'.

show abstract

Toward Earth system modeling with resolved clouds and ocean submesoscales on heterogeneous many-core HPCs

Cited by 10 publications

References 88 publications

The Parallel Performance of SLNE Atmosphere-Ocean-Sea Ice Coupled Model

The Parallel Performance of SLNE Atmosphere-Ocean-Sea Ice Coupled Model

Observations Reveal Intense Air‐Sea Exchanges Over Submesoscale Ocean Front

From theory to RAPID AMOC observations: a personal voyage of discovery

Contact Info

Product

Resources

About